Abstract Introduction/Objective After professional transcription service is eliminated, pathologists inevitably undertake the task of diagnostic data entry into pathology repot by adapting a variety of methods such as speech recognition, manual typing, and pre-texted command. Errors and inefficiency in reporting remain common problems, especially for information with unusual syntax such as genotype or nucleotide sequences. To overcome these shortcomings, we introduce here a novel application of a well-established technology as a complementary method, namely 2- dimensional (2D) barcode symbology. Methods/Case Report Commonly used diagnostic wordings of pathology reports including specimen type, surgical procedure, diagnosis, and test results are collated and organized by organ (specimen type) and by their frequency of usage/occurrence. Next, 2D data matrix barcodes are created for these diagnostic wordings using a on-line tool (www.free-barcode-generator.net/datamatrix/). The 2D barcodes along with their text are displayed on the computer screen (or printed out as a booklet). A 2D barcode scanner (Symbol LS2208, Motorola) was used to retrieve the text information from the barcodes and transfer into the pathology report. To assess the efficacy of this barcode method, we evaluated the time of data entry into reports for 117 routine cases using an on-line stopwatch and compared with those by other data entry methods. Results (if a Case Study enter NA) Unlike manual typing or speech recognition, the barcode method did not introduce typographic or phonosemantic errors since the method simply transferred pre-texted and proof-read text content to report. It was also faster than manual typing or speech recognition, and its speed was comparable to that of the pre-text method integrated in LIS but did not require human memorization of innumerable text commands to retrieve desired diagnosis wordings. Conclusion Our preliminary results demonstrated that the diagnostic data entry time was reduced from 28.5% by other methods to 22.1% by the barcode method although due to the small sample size, statistical analysis was not conclusive.
Read full abstract